Skip to main content

Inspecting Live Resources

Quick-reference for agents investigating issues across the Cloud Control Plane infrastructure.

Environment Quick Reference

EnvAWS AccountRegionDefault CLI Profile
Staging468036072962us-east-1Default
Preprod010928202142us-east-1Requires --profile
Prod023568249301us-east-1Requires --profile

Cloudflare accounts: 3a8e992c9f607dcb3b401878264df92e (dev/staging/preprod), 96d9ae8ad2fe71a3c6ead54f3a775c05 (prod).

Naming Convention

Most AWS resources follow {env_prefix}-{ResourceName}:

  • Dev: dev-{branch}- (e.g., dev-oliver-EcomIndexSettingsTable)
  • Staging: staging-
  • Preprod: preprod-
  • Prod: prod-

Exception: UsersAccountsTable has no prefix in staging.

End-to-End Flow Guides

Trace a customer action through every service hop, with what to check at each step:

FlowDescription
SearchSearch query through proxy → KV → Marqo and back
Add DocumentsProduct sync from Shopify → S3 → SQS → Indexer → Marqo
RecommendationsSimilar, suggestions, and personalized recommendations
Agentic SearchAI search with Gemini, caching, conversations
MerchandisingRules from Admin/Controller → DDB → exporter → KV → search runtime
SignupRegistration through Cognito, Stripe, DynamoDB
LoginAuthentication and API authorization
Settings SyncDynamoDB → settings exporter → Cloudflare KV

Start Here: Which Component?

If the issue involves...Start with
Search queries, results, latencySearch Proxy
Merchandising rules, pins, boosts, filters, recencyMerchandising
AI/conversational searchAgentic Search
Product indexing, Shopify syncEcommerce
Admin dashboard, query configs, forksAdmin
User signup, login, accounts, billingController
Console web app, monolith APIControl Plane

Resource Type Guides

How to inspect each type of resource with CLI commands:

ResourceGuide
DynamoDB tablesresources/dynamodb.md
DynamoDB access patterns (cheat sheet)ddb-access-patterns.md
Lambda functionsresources/lambda.md
API Gatewayresources/api-gateway.md
Cloudflare Workers / KV / DOresources/cloudflare-workers.md
CloudWatch (logs, alarms, dashboards)resources/cloudwatch.md
Cognito (user pools, auth)resources/cognito.md
S3 bucketsresources/s3.md
SQS queuesresources/sqs.md
Secrets Managerresources/secrets-manager.md
ECS / Fargateresources/ecs.md
Elastic Beanstalkresources/elastic-beanstalk.md
WAFresources/waf.md

Data Plane Cell Gateways

IAM-authenticated API Gateway endpoints per data plane cell. Config lives in infra/*/config/{env}.json under data_plane_cells.

CellGateway IDAccount
S (staging)n6wwdwmk2m468036072962
PP (preprod)me25xyukr9010928202142
P1 (prod)m6j1teqnte023568249301

Invoke with IAM auth:

python3 -c "
import requests, json
from aws_requests_auth.boto_utils import BotoAWSRequestsAuth
host = 'n6wwdwmk2m.execute-api.us-east-1.amazonaws.com'
auth = BotoAWSRequestsAuth(aws_host=host, aws_region='us-east-1', aws_service='execute-api')
resp = requests.post(f'https://{host}/prod/v2/indexes/list', auth=auth,
json={'systemAccountId': 'fnqm9psx'})
print(json.dumps(resp.json(), indent=2))
"

Common Investigation Patterns

"Why is search returning wrong/stale results?"

  1. Check index settings in DDB: DynamoDB -> EcomIndexSettingsTable
  2. Check KV cache: Cloudflare Workers
  3. Check settings exporter: Lambda -> EcomSettingsExporterLambda
  4. Check search proxy logs: Cloudflare Workers -> {env}-ecom-api

"Why is indexing stuck?"

  1. Check indexer jobs: DynamoDB -> EcomIndexerJobsTable
  2. Check indexer Lambda: Lambda -> EcomIndexerFunction
  3. Check SQS queue depth: SQS
  4. Check product data: S3 -> ecom-product-data-bucket

"Why can't a user sign in?"

  1. Check Cognito user: Cognito
  2. Check auth Lambda logs: Lambda -> ControllerAuthLambda-{env}
  3. Check controller health: Elastic Beanstalk

"Why is the admin dashboard broken?"

  1. Check admin worker: Cloudflare Workers -> {env}-admin-api
  2. Check admin Lambda: Lambda -> AdminLambda
  3. Check API Gateway: API Gateway -> AdminInternalApi

"Why are alarms firing?"

  1. Check alarm state: CloudWatch
  2. Follow the alarm to its source metric/resource
  3. Check associated Lambda/service logs

Infrastructure Code Locations

ComponentCDK/PulumiConfig
Control Plane (Console)infra/aws/stacks/infra/aws/config/
Ecommerceinfra/ecom/stacks/infra/ecom/config/
Admininfra/admin/stacks/infra/admin/config/
Controllerinfra/controller/stacks/infra/controller/configurations/
Cloudflare Workerscomponents/{worker}/wrangler.tomlPer-env sections in wrangler.toml